One of the most important cellular functions of integrin-mediated cell adhesion to extracellular matrix (ECM) is to promote cell survival by mediating anti-apoptotic signals from ECM to cells. Most types of cells depend on ECM attachment for survival; if denied it, they undergo apoptosis (anoikis). Malignant cells are less dependent on integrin-mediated survival signals than normal cells. A pathway activated by some, but not all, integrins up-regulates the anti-apoptotic protein Bcl-2. In screening for cDNAs that regulate this pathway, we isolated a eDNA for a novel protein, which we have named Bit1. Bit1 is a mitochondrial protein which, when released into the cytoplasm, induces apoptosis. Yeast two-hybrid screening and biochemical data revealed an interaction between Bit1 and the transcriptional regulator protein Amino-terminal Enhancer of Split (AES). Forced expression of AES also causes apoptosis. The degree of apoptosis correlates with the level of Bit1-AES complexes in the cytoplasm, suggesting that the complex is the pro-apoptotic factor. Transfection with Bcl-2, Bcl-xL, various caspase inhibitors, or activated PI3-K, Akt or H-Ras, does not block apoptosis induced by AES or cytoplasmic Bit1. Plating cells onto fibronectin is the only treatment identified so far that counteracts the apoptosis-inducing effect of Bit1 and AES. The use of alternative adhesion surfaces and antibody inhibition experiments suggests that the anti-apoptotic effect of Bit1/AES is regulated by specific integrins. The Bit1/AES pathway may be, at least in part, responsible for the anti-apoptotic effect of integrin-mediated cell adhesion. This application proposes experiments to establish the integrin-regulation of the Bit1/AES pathway and to identify the subset of integrins that regulate it, The integrin regulation studies will focus on the formation of the Bit1/AES complex, the presumed pro-apoptotic factor. A crystallographic study on the complex is also proposed. Finally, we will analyze the in vivo function of Bit1 in normal physiology and in tumorigenesis by using a conditional Bit1 gene knockout, which is already at hand. The results of these studies may delineate a signaling pathway that is of fundamental importance in the anchorage dependence of normal cells. Cells that become malignant may bypass this pathway in becoming anchorage independent and metastatic.